Background: Activities of Daily Living (ADLs) are essential tasks performed at home and used in healthcare to monitor sedentary behavior, track rehabilitation therapy, and monitor chronic obstructive pulmonary disease. The Barthel Index, used by healthcare professionals, has limitations due to its subjectivity. Human activity recognition (HAR) is a more accurate method using Information and Communication Technologies (ICTs) to assess ADLs more accurately. This work aims to create a singular, adaptable, and heterogeneous ADL dataset that integrates information from various sources, ensuring a rich representation of different individuals and environments. Methods: A literature review was conducted in Scopus, the University of California Irvine (UCI) Machine Learning Repository, Google Dataset Search, and the University of Cauca Repository to obtain datasets related to ADLs. Inclusion criteria were defined, and a list of dataset characteristics was made to integrate multiple datasets. Twenty-nine datasets were identified, including data from various accelerometers, gyroscopes, inclinometers, and heart rate monitors. These datasets were classified and analyzed from the review. Tasks such as dataset selection, categorization, analysis, cleaning, normalization, and data integration were performed. Results: The resulting unified dataset contained 238,990 samples, 56 activities, and 52 columns. The integrated dataset features a wealth of information from diverse individuals and environments, improving its adaptability for various applications. Conclusions: In particular, it can be used in various data science projects related to ADL and HAR, and due to the integration of diverse data sources, it is potentially useful in addressing bias in and improving the generalizability of machine learning models.
- Publication type
- Journal Article MeSH
Background/Objectives: Activities of Daily Living (ADLs) are crucial for assessing an individual's autonomy, encompassing tasks such as eating, dressing, and moving around, among others. Predicting these activities is part of health monitoring, elderly care, and intelligent systems, improving quality of life, and facilitating early dependency detection, all of which are relevant components of personalized health and social care. However, the automatic classification of ADLs from sensor data remains challenging due to high variability in human behavior, sensor noise, and discrepancies in data acquisition protocols. These challenges limit the accuracy and applicability of existing solutions. This study details the modeling and evaluation of real-time ADL classification models based on batch learning (BL) and stream learning (SL) algorithms. Methods: The methodology followed is the Cross-Industry Standard Process for Data Mining (CRISP-DM). The models were trained with a comprehensive dataset integrating 23 ADL-centric datasets using accelerometers and gyroscopes data. The data were preprocessed by applying normalization and sampling rate unification techniques, and finally, relevant sensor locations on the body were selected. Results: After cleaning and debugging, a final dataset was generated, containing 238,990 samples, 56 activities, and 52 columns. The study compared models trained with BL and SL algorithms, evaluating their performance under various classification scenarios using accuracy, area under the curve (AUC), and F1-score metrics. Finally, a mobile application was developed to classify ADLs in real time (feeding data from a dataset). Conclusions: The outcome of this study can be used in various data science projects related to ADL and Human activity recognition (HAR), and due to the integration of diverse data sources, it is potentially useful to address bias and improve generalizability in Machine Learning models. The principal advantage of online learning algorithms is dynamically adapting to data changes, representing a significant advance in personal autonomy and health care monitoring.
- Publication type
- Journal Article MeSH
Precise localization of the epileptogenic zone is pivotal for planning minimally invasive surgeries in drug-resistant epilepsy. Here, we present a graph neural network (GNN) framework that integrates interictal intracranial EEG features, electrode topology, and MRI features to automate epilepsy surgery planning. We retrospectively evaluated the model using leave-one-patient-out cross-validation on a dataset of 80 drug-resistant epilepsy patients treated at St. Anne's University Hospital (Brno, Czech Republic), comprising 31 patients with good postsurgical outcomes (Engel I) and 49 with poor outcomes (Engel II-IV). The GNN predictions demonstrated a significantly better (P < 0.05, Mann-Whitney-U test) area under the precision-recall curve in patients with good outcomes (area under the precision-recall curve: 0.69) compared with those with poor outcomes (area under the precision-recall curve: 0.33), indicating that the model captures clinically relevant targets in successful cases. In patients with poor outcomes, the graph neural network proposed alternative intervention sites that diverged from the original clinical plans, highlighting its potential to identify alternative therapeutic targets. We show that topology-aware GNNs significantly outperformed (P < 0.05, Wilcoxon signed-rank test) traditional neural networks while using the same intracranial EEG features, emphasizing the importance of incorporating implantation topology into predictive models. These findings uncover the potential of GNNs to automatically suggest targets for epilepsy surgery, which can assist the clinical team during the planning process.
- Publication type
- Journal Article MeSH
Protein misfolding diseases, including α1-antitrypsin deficiency (AATD), pose substantial health challenges, with their cellular progression still poorly understood1-3. We use spatial proteomics by mass spectrometry and machine learning to map AATD in human liver tissue. Combining Deep Visual Proteomics (DVP) with single-cell analysis4,5, we probe intact patient biopsies to resolve molecular events during hepatocyte stress in pseudotime across fibrosis stages. We achieve proteome depth of up to 4,300 proteins from one-third of a single cell in formalin-fixed, paraffin-embedded tissue. This dataset reveals a potentially clinically actionable peroxisomal upregulation that precedes the canonical unfolded protein response. Our single-cell proteomics data show α1-antitrypsin accumulation is largely cell-intrinsic, with minimal stress propagation between hepatocytes. We integrated proteomic data with artificial intelligence-guided image-based phenotyping across several disease stages, revealing a late-stage hepatocyte phenotype characterized by globular protein aggregates and distinct proteomic signatures, notably including elevated TNFSF10 (also known as TRAIL) amounts. This phenotype may represent a critical disease progression stage. Our study offers new insights into AATD pathogenesis and introduces a powerful methodology for high-resolution, in situ proteomic analysis of complex tissues. This approach holds potential to unravel molecular mechanisms in various protein misfolding disorders, setting a new standard for understanding disease progression at the single-cell level in human tissue.
- MeSH
- alpha 1-Antitrypsin metabolism MeSH
- Single-Cell Analysis MeSH
- alpha 1-Antitrypsin Deficiency * pathology metabolism genetics MeSH
- Phenotype MeSH
- Hepatocytes metabolism pathology MeSH
- Liver Cirrhosis pathology metabolism MeSH
- Liver pathology metabolism MeSH
- Humans MeSH
- Disease Progression MeSH
- Proteome * analysis metabolism MeSH
- Proteomics * methods MeSH
- Unfolded Protein Response MeSH
- Machine Learning MeSH
- Check Tag
- Humans MeSH
- Male MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
BACKGROUND AND HYPOTHESIS: Cognitive impairments are a core feature of psychosis that are often evident before illness onset and have substantial impact on both clinical and real-world functional outcomes. Therefore, these are an excellent target for stratification and early detection in order to facilitate early intervention. While many studies have aimed to characterize the effects of cognition at the group level and others have aimed to detect individual differences by referencing subjects against existing norms, these studies have limited generalizability across clinical populations, demographic backgrounds, and instruments and do not fully account for the interindividual heterogeneity inherent in psychosis. STUDY DESIGN: Here, we outline the rationale, design, and analysis plan for the PRECOGNITION project, which aims to address these challenges. STUDY RESULTS: This project is a collaboration between partners in 5 European countries. The project will not generate any primary data, but by leveraging existing datasets and combining these with novel analytic methods, it will produce multiple contributions including: (i) translating normative modeling approaches pioneered in brain imaging to psychosis data, to yield "cognitive growth charts" for longitudinal tracking and individual prediction; (ii) developing machine learning models for harmonizing and stratifying cohorts on the basis of these models; and (iii) providing integrated next-generation norms, having broad sociodemographic coverage including different languages and distinct norms for individuals with psychosis and unaffected individuals. CONCLUSIONS: This study will enable precision stratification of psychosis cohorts and furnish predictions for a broad range of functional outcome measures. It will be guided throughout by lived experience experts.
- Publication type
- Journal Article MeSH
BACKGROUND: Advancements in artificial intelligence (AI) and machine learning (ML) have revolutionized the medical field and transformed translational medicine. These technologies enable more accurate disease trajectory models while enhancing patient-centered care. However, challenges such as heterogeneous datasets, class imbalance, and scalability remain barriers to achieving optimal predictive performance. METHODS: This study proposes a novel AI-based framework that integrates Gradient Boosting Machines (GBM) and Deep Neural Networks (DNN) to address these challenges. The framework was evaluated using two distinct datasets: MIMIC-IV, a critical care database containing clinical data of critically ill patients, and the UK Biobank, which comprises genetic, clinical, and lifestyle data from 500,000 participants. Key performance metrics, including Accuracy, Precision, Recall, F1-Score, and AUROC, were used to assess the framework against traditional and advanced ML models. RESULTS: The proposed framework demonstrated superior performance compared to classical models such as Logistic Regression, Random Forest, Support Vector Machines (SVM), and Neural Networks. For example, on the UK Biobank dataset, the model achieved an AUROC of 0.96, significantly outperforming Neural Networks (0.92). The framework was also efficient, requiring only 32.4 s for training on MIMIC-IV, with low prediction latency, making it suitable for real-time applications. CONCLUSIONS: The proposed AI-based framework effectively addresses critical challenges in translational medicine, offering superior predictive accuracy and efficiency. Its robust performance across diverse datasets highlights its potential for integration into real-time clinical decision support systems, facilitating personalized medicine and improving patient outcomes. Future research will focus on enhancing scalability and interpretability for broader clinical applications.
- MeSH
- Databases, Factual MeSH
- Humans MeSH
- Neural Networks, Computer MeSH
- Patient-Centered Care * MeSH
- Machine Learning * MeSH
- Translational Science, Biomedical MeSH
- Translational Research, Biomedical MeSH
- Artificial Intelligence * MeSH
- Treatment Outcome MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
BACKGROUND & AIMS: Exogenous recombinant fibroblast growth factor 20 (FGF20) protein has been proved to treat ulcerative colitis; however, its mechanism of action remains unclear. This study aimed to explore the role and mechanism of action of FGF20 in ulcerative colitis. METHODS: Data from patients with ulcerative colitis were analyzed using the Gene Expression Omnibus dataset. A murine colitis model was established by administering 2% dextran sodium sulfate. FGF20 knockout mice and Adenoassociated viruses (AAV)-FGF20-treated mice were used to elucidate the specific mechanisms. Proteomic analysis was conducted to identify differentially expressed genes. RESULTS: FGF20 levels were significantly elevated in the colonic tissues of subjects and mice with colitis. FGF20 deficiency exacerbated dextran sodium sulfate-induced colitis; in contrast, FGF20 replenishment alleviated colitis through 2 principal mechanisms: restoration of impaired intestinal epithelial barrier integrity, and inhibition of M1 macrophage polarization. Notably, S100A9 was identified as a pivotal downstream target of FGF20, which was further demonstrated by pharmacologic inhibition and overexpression experiments of S100A9 using paquinimod (a specific inhibitor of S100A9) and AAV-S100A9 in FGF20 knockout and AAV-FGF20 mice with colitis, respectively. Additionally, the nuclear factor-κB pathway was found to be involved in the process by which FGF20 regulates S100A9 to counteract colitis. CONCLUSIONS: These results suggest that FGF20 acts as a negative regulator of S100A9 and nuclear factor-κB, thereby inhibiting M1 macrophage polarization and restoring intestinal epithelial barrier integrity in mice with dextran sodium sulfate-induced colitis. FGF20 may serve as a potential therapeutic target for the treatment of ulcerative colitis.
- MeSH
- Fibroblast Growth Factors * metabolism genetics pharmacology MeSH
- Calgranulin B * metabolism genetics MeSH
- Colitis * chemically induced MeSH
- Humans MeSH
- Macrophages * immunology metabolism drug effects MeSH
- Disease Models, Animal MeSH
- Mice, Inbred C57BL MeSH
- Mice, Knockout MeSH
- Mice MeSH
- NF-kappa B * metabolism MeSH
- Signal Transduction MeSH
- Dextran Sulfate toxicity MeSH
- Intestinal Mucosa * pathology metabolism immunology drug effects MeSH
- Colitis, Ulcerative * pathology chemically induced immunology metabolism drug therapy MeSH
- Animals MeSH
- Check Tag
- Humans MeSH
- Male MeSH
- Mice MeSH
- Animals MeSH
- Publication type
- Journal Article MeSH
BACKGROUND: Early detection of colorectal cancer (CRC) significantly improves its management and patients' survival. Circular RNAs (circRNAs) are peculiar covalently closed transcripts involved in gene expression modulation whose dysregulation has been extensively reported in CRC cells. However, little is known about their alterations in the early phases of colorectal carcinogenesis. METHODS: In this study, we performed an integrative analysis of circRNA profiles in RNA-sequencing (RNA-Seq) data of 96 colorectal cancers, 27 adenomas, and matched adjacent mucosa tissues. We also investigated the levels of cognate linear transcripts and those of regulating RNA-binding proteins (RBPs). Levels of circRNA-interacting microRNAs (miRNAs) were explored by integrating data of small RNA-Seq performed on the same samples. RESULTS: Our results revealed a significant dysregulation of 34 circRNAs (paired adj. p < 0.05), almost exclusively downregulated in tumor tissues and, prevalently, in early disease stages. This downregulation was associated with decreased expression of circRNA host genes and those encoding for RBPs involved in circRNA biogenesis, including NOVA1, RBMS3, and MBNL1. Guilt-by-association analysis showed that dysregulated circRNAs correlated with increased predicted activity of cell proliferation, DNA repair, and c-Myc signaling pathways. Functional analysis showed interactions among dysregulated circRNAs, RBPs, and miRNAs, which were supported by significant correlations among their expression levels. Findings were validated in independent cohorts and public datasets, and the downregulation of circLPAR1(2,3) and circLINC00632(5) was validated by ddPCR. CONCLUSIONS: These results support that multiple altered regulatory mechanisms may contribute to the reduction of circRNA levels that characterize early colorectal carcinogenesis.
- Publication type
- Journal Article MeSH
BACKGROUND AND OBJECTIVE: Metabolomic interaction networks provide critical insights into the dynamic relationships between metabolites and their regulatory mechanisms. This study introduces MInfer, a novel computational framework that integrates outputs from MetaboAnalyst, a widely used metabolomic analysis tool, with Jacobian analysis to enhance the derivation and interpretation of these networks. METHODS: MInfer combines the comprehensive data processing capabilities of MetaboAnalyst with the mathematical modeling power of Jacobian analysis. This framework was applied to various metabolomic datasets, employing advanced statistical tests to construct interaction networks and identify key metabolic pathways. RESULTS: The application of MInfer revealed significant metabolic pathways and potential regulatory mechanisms across multiple datasets. The framework demonstrated high precision, sensitivity, and specificity in identifying interactions, enabling robust network interpretations. CONCLUSIONS: MInfer enhances the interpretation of metabolomic data by providing detailed interaction networks and uncovering key regulatory insights. This tool holds significant potential for advancing the study of complex biological systems.
- MeSH
- Algorithms MeSH
- Humans MeSH
- Metabolic Networks and Pathways * MeSH
- Metabolomics * MeSH
- Software MeSH
- Computational Biology MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
Rare diseases may affect the quality of life of patients and be life-threatening. Therapeutic opportunities are often limited, in part because of the lack of understanding of the molecular mechanisms underlying these diseases. This can be ascribed to the low prevalence of rare diseases and therefore the lower sample sizes available for research. A way to overcome this is to integrate experimental rare disease data with prior knowledge using network-based methods. Taking this one step further, we hypothesized that combining and analyzing the results from multiple network-based methods could provide data-driven hypotheses of pathogenic mechanisms from multiple perspectives.We analyzed a Huntington's disease transcriptomics dataset using six network-based methods in a collaborative way. These methods either inherently reported enriched annotation terms or their results were fed into enrichment analyses. The resulting significantly enriched Reactome pathways were then summarized using the ontological hierarchy which allowed the integration and interpretation of outputs from multiple methods. Among the resulting enriched pathways, there are pathways that have been shown previously to be involved in Huntington's disease and pathways whose direct contribution to disease pathogenesis remains unclear and requires further investigation.In summary, our study shows that collaborative network analysis approaches are well-suited to study rare diseases, as they provide hypotheses for pathogenic mechanisms from multiple perspectives. Applying different methods to the same case study can uncover different disease mechanisms that would not be apparent with the application of a single method.
